Proc. Natl. Acad. Sci. USAVol. 87, pp. 6777-6780, September 1990Neurobiology

Changed distribution of sodium channels along demyelinated axons(ion channels/neural conduction/myelin/antibodies/doxorubicin)

JOHN D. ENGLAND*t, FABIA GAMBONI*, S. ROCK LEVINSON*, AND THOMAS E. FINGER§Departments of *Neurology, tPhysiology, and §Cellular and Structural Biology, University of Colorado Health Sciences Center, Denver, CO 80262

Communicated by Theodore H. Bullock, June 22, 1990

ABSTRACT Voltage-gated sodium channels are largelylocalized to the nodes of Ranvier in myelinated axons, provid-ing a physiological basis for saltatory conduction. What hap-pens to these channels in demyelinated axons is not known withcertainty. Experimentally demyelinated axons were examinedby using a well-characterized polyclonal antibody directedagainst sodium channels. Immunocytochemical and radioim-munoassay data were consistent with the distribution of anincreased number of sodium channels along segments of pre-viously internodal axon. These findings affirm the plasticity ofsodium channels in demyelinated axolemma and may be rele-vant to understanding how axons recover conduction afterdemyelination.

The sodium channel is a transmembrane protein that medi-ates the voltage-dependent sodium permeability of electri-cally excitable membranes. The presence ofsodium channelsis of obvious importance for the generation and propagationof action potentials along axolemma. In normal myelinatedaxons sodium channels are largely localized to the nodes ofRanvier. The most recent electrophysiological and biochem-ical studies demonstrate a sodium channel density of severalthousand channels per gm2 at the nodes ofRanvier comparedwith a density of

6778 Neurobiology: England et al.

postfixed with 2% (wt/vol) Os04/0.1 M phosphate buffer,dehydrated in a series of graded ethanol, passed throughpropylene oxide, and embedded in Epon 812. For electronmicroscopy ultrathin sections were cut, doubly stained withuranyl acetate and lead citrate and examined with a Philips300 electron microscope.Radioimmunoassay. The pooled nerve preparations were

homogenized in 1% Lubrol PX and centrifuged at 45,700 x gfor 25 min. The supernatants were diluted in 0.1 M sodiumphosphate buffer, pH = 7.5/0.1% bovine serum albumin/1%Lubrol PX to a concentration of

Proc. Natl. Acad. Sci. USA 87 (1990) 6779

m.. .'. '. 7--l-. - -.' -,-i- i,:-% -" 7,:.

.. -T

e.", 'f

1W ,~~~FIG. 2. Electron micrograph of an adjacent section of the nerve shown in Fig. 1C. Axons (Ax) are surrounded only by a Schwann cell basal

lamina (arrowhead) without Schwann cell cytoplasm or myelin. (Bar = 1 am.)

seen only in restricted regions corresponding in dimension tonodal axolemma (i.e., -1 A&m in length).Immunocytochemistry of Demyelinated Nerve. Thirty of the

doxorubicin-injected nerves were examined immunohis-tochemically with anti-TTXR antibody at various times afterinjection. As early as 14-21 days after injection specificimmunoreactivity occurred over comparatively long seg-ments of demyelinated axons (Fig. 1 B and C). Thesesegments ofimmunoreactivity ranged in length from 35 Aum to72 um-several times longer than nodes of Ranvier. Somedemyelinated axons showed multiple regions of specificimmunoreactivity, each extending over relatively longstretches of bare axolemma. This kind of immunofluores-cence was not observed in either previously adsorbed anti-TTXR antibody preparations or in preparations incubatedwith normal rabbit serum instead of anti-TTXR antibody.

Electron Microscopic Observations. Correlative electronmicroscopic studies confirmed that doxorubicin had pro-duced demyelination in these nerves (Fig. 2) just as it does inperipheral nerve of rat (14). Several longitudinal sections ofaxons from the same nerves and adjacent to the axonsexamined immunocytochemically showed long segments ofdemyelinated axons surrounded only by basal lamina.Radioimmunoassay. The same anti-TTXR antibody was

utilized in a radioimmunoassay (RIA) comparing the quantityof sodium channels in control (myelinated) nerves and ex-perimental (demyelinated) nerves. The demyelinated nerveswere assayed at day 14 after injection of doxorubicin so thatthese results could be directly correlated with the immuno-cytochemical data. Fifty-nine experimental nerves were par-titioned into four pooled nerve preparations; 72 controlnerves were partitioned into seven pooled nerve prepara-tions. This RIA indicated a highly significant (P < 0.01,2-tailed Student's t test) 4-fold increase in the number ofsodium channels per unit of wet weight in the experimental(demyelinated) nerves as compared with the control (myeli-nated) nerves (mean sodium channel concentration per wetweight of nerve = 18.328 pM per g + SD of 5.288 for theexperimental group vs. 4.473 pM per g ± SD of 1.073 for thecontrol group).

DISCUSSIONThese experiments indicate that sodium channels form alongthe internodal segments of demyelinated axons. Moreover,the intensity and length ofthe immunoreactivity coupled withthe RIA data suggest that channels are being placed de novoin these locations. The source of these additional channels isan important question. A sizable body of evidence suggeststhat some Schwann cells contain voltage-gated sodium chan-nels (15-17), leading to the hypothesis that axonal sodium

channels could be locally manufactured within these cells(17). In the present study the appearance of sodium channelsoccurred along axons without nearby Schwann cells, imply-ing that the channels were synthesized and inserted whollywithin neuron-axons. Thus, if Schwann cells are involved inchannel turnover or maintenance for axons, they probablyprovide only an ancillary source. Other recent work (13)using similar immunocytochemical techniques combinedwith electrophysiological methods has demonstrated thatsodium channels accumulate only at the hyperexcitable prox-imal endings of ligated axons (neuromas). This and otherstudies (18) suggest that new sodium channel distributionscan arise from neuron-axonal influences alone without glialparticipation.These kinds of studies are one step in understanding how

axons recover function after demyelination. Remyelinationand the re-establishment of saltatory conduction are a majormeans by which conduction can be restored. Discrete foci ofinward membrane current, presumably representing futurenodes of Ranvier, have been found to precede remyelinationin axons demyelinated with lysophosphatidyl choline (19).This observation, by itself, would seem to indicate theformation of new aggregates of sodium channels in demyeli-nated axons. When remyelination occurs, internodal dis-tances are shorter, indicating the formation of nodes ofRanvier in previously internodal axolemma. In addition,pharmacological studies in remyelinated peripheral nervehave shown an increase in saxitoxin binding that is propor-tional to this increase in nodal area (20); these results provideevidence that recently formed nodes along remyelinatedaxons have a relatively normal density of sodium channels,presumably reflecting channel synthesis and insertion afterdemyelination.Another possible mechanism of axonal recovery after

demyelination is continuous conduction. This mechanismcould be particularly important in recovery from demyelina-tion in the central nervous system (e.g., in multiple sclerosis),where limited remyelination occurs (21). Bostock and Sears(22, 23) demonstrated continuous conduction along shortsegments of axons previously demyelinated by diphtheriatoxin, again suggesting a reorganization of sodium channels.Our findings provide direct morphological and immuno-

logical characterization of sodium channel changes in demy-elinated axons. However, whether the remodeled sodiumchannels seen in our study support continuous conduction orserve simply as a prelude to the formation of new nodes ofRanvier is not yet known. Extension of these studies shouldhelp us understand how axons can recover function afterdemyelinative insults. This area of inquiry is particularlyimportant because the primary effect of several human dis-

Neurobiology: England et al.

Dow

nloa

ded

by g

uest

on

Apr

il 7,

202

1

6780 Neurobiology: England et al.

eases is demyelination of either the peripheral or centralnervous system.

This work was supported by National Institutes of Health GrantsNS15879 (S.R.L.) and DC00244 (T.E.F.). J.D.E. is supported by aClinical Investigator Development Award from the National Instituteof Neurological Disorders and Stroke (NS01272-03).

1. Ritchie, J. M. & Rogart, R. B. (1977) Proc. Nadl. Acad. Sci.USA 74, 211-215.

2. Chiu, S. Y. (1980) J. Physiol. (London) 309, 499-519.3. Neumcke, B. & Stampfli, R. (1982) J. Physiol. (London) 329,

163-184.4. Ellisman, M. H. & Levinson, S. R. (1982) Proc. Nati. Acad.

Sci. USA 79, 6707-6711.5. Waxman, S. G. & Ritchie, J. M. (1985) Science 228,1502-1507.6. Smith, K. J. & Hall, S. M. (1980) J. Neurol. Sci. 48, 201-219.7. Sumner, A. J. (1981) Ann. Neurol. 9, Suppl., 28-30.8. Albers, J. W., Donofrio, P. D. & 'McGonagle, T. K. (1985)

Muscle Nerve 8, 528-539.9. Miller, J. A., Agnew, W. S. & Levinson, S. R. (1983) Biochem-

istry 22, 462-470.10.' Duch, D. S. & Levinson, S. R. (1987) J. Membr. Biol. 98,

43-55.

11. Recio-Pinto, E., Duch, D. S., Levinson, S. R. & Urban, B. W.(1987) J. Gen. Physiol. 90, 375-395.

12. Thornhill, W. B. & Levinson, S. R. (1987) Biochemistry 26,4381-4388.

13. Devor, M., Keller, C. H., Deerinck, T. J., Levinson, S. R. &Ellisman, M. H. (1989) Neurosci. Lett. 102, 149-154.

14. Englanrd, J. D., Rhee, E. K., Said,. G. & Sumner, A. J. (1988)Brain 111, 901-913.

15. Ritchie, J. M. & Rang, H. P. (1983) Proc. Natl. Acad. Sci. USA80, 2803-2807.

16. Chiu, S. Y., Shrager, P. & Ritchie, J. M. (1984) Nature (Lon-don) 311, 156-157.

17. Shrager, P., Chiu, S. Y. & Ritchie, J. M. (1985) Proc. Natl.Acad. Sci. USA 82, 948-952.

18. Brismar, T. & Gilly, W. F. (1987) Proc. Natl. Acad. Sci. USA84, 1459-1463.

19. Smith, K. J., Bostock, H. & Hall, S. M. (1982) J. Neurol. Sci.54, 13-31.

20. Ritchie, J. M., Rang, H. P. & Pellegrino, R. (1981) Nature(London) 294, 257-259.

21. McDon-ald, W. I. (1974) Br. Med. Bull. 30, 186-189.22. Bostock, H. & Sears, T. A. -(1976) Nature (London) 263,

7g6-787.23. Bostock, H. & Sears, T. A. (1978) J. Physiol. (London) 280,

273-301.

Proc. Natl. Acad. Sci. USA 87 (1990)

Dow

nloa

ded

by g

uest

on

Apr

il 7,

202

1